Macrophage inflammatory proteins MIP-1 and MIP-2 are ... - CiteSeerX
Macrophage inflammatory involved in T cell-mediated
proteins MIP-1 and MIP-2 neutrophil recruitment
are
Rui Appelberg Centro
de Cilologia
Institute,
Abstract:
Mice
Mycobacterium
University
Experimental
(Instituto
of Porto,
Nacional
de Investigacdo
intraperitoneally
(i.p.)
infected
with
formed
bovis
bacille Calmette-Gu#{233}rin (BCG) respond to an i.p. challenge with mycobacterial antigen with an acute and extensive accumulation of neutrophils. This influx was not mimicked by the inoculation of recombinant interferon--y (IFN-’y) or tumor necrosis factor a (TNF-a). The antigen-induced recruitment of neutrophils was not affected by coinoculation of anti-IFN-y antibodies, was enhanced by anti-TNF-a antisera, and was significantly reduced by antisera against macrophage inflammatory proteins MIP-1 and MIP-2. The latter two sera had no additive effects. We hypothesize that mycobacteria-specific T cells are triggered by antigen to secrete MIP-1 and MIP-2, which directly mediate, at least partly, the influx of neutrophils that takes place in
J.
this
model.
Key
Words: factor
necrosis
Leukoc. neutrophils, macrophage
.
Biol.
52:
303-306;
T cell-mediated inflammatory
. interferon-n1 proteins
Hemacolor
macrophage
inflammatory
a generous New York).
gift
MIP-2
were
University,
Bacteria
Sciences
(Detroit,
preparations
from
Merck
(Darmstadt,
and
bacterial
from
antigen
MIP-1 and (Rockefeller
proteins
Dr.
B. Sherry
preparations
bovis BCG (TMCC 1011, strain Pasteur) and (ATCC 25291) were grown in Middlebrook 7H9 supplemented with 0.04% Tween 80. Inocula were from mid-log phase cultures by pelleting the bacresuspending in a small volume of saline with 0.04% 80, briefly sonicating the bacterial suspension, and
Mycobacterium
freezing at -70#{176}C until use. Listeria monocytogenes (strain EGD) was grown in Antibiotic 3 broth. Cytosolic proteins from M. bovis BCG, M. avium, and L. monocytogenes were prepared as described [1] and used as the antigen sources.
Animals Female C57BL/6 Institute (Oeiras, hygiene conditions, ad libitum.
Priming
and
mice
were Portugal), and fed
challenge
MI).
The
was
per-
intraperitoneal; necrosis
Reprint
from under and
the
Gulbenkian conventional acidified water
of mice
Received
of
bacille
Calmette-Gu#{233}rin; unit;
MIP-I,
IFN-y,
macrophage
peptide
1;
BSA,
interferon-7;
serum
IL-I,
inflammatory
PBS,
bovine protein
phosphate-buffered
albu-
interleukin-I; I;
saline;
NAP-i,
TNF,
factor.
requests:
do Campo
Journal
purchased kept food pellets
Mice were inoculated intraperitoneally (i.p.) with 106 colonyforming units (CFU) of BCG and 3 weeks later were administered the antigen or cytokine stimulus by the same route. Five hours after the second inoculation, the animals were sacrificed and the peritoneal exudate studied as described elsewhere [1, 17]. The results are expressed as the arithmetic means of neutrophils per peritoneal cavity from four mice per time point ± 1 standard deviation of the mean. The statistical analysis was done with the Statworks program using Student’s I-test.
tumor
Difco
kit
antimouse
neutrophil-activating
were from of cytospin
Biomedical
serum albumin (BSA) was from Sigma Recombinant mouse IFN-’y, recombinant mouse TNF-a, and rabbit antimouse TNF-a polyclonal serum were purchased from Genzyme (Boston, MA). Antimouse IFN--y monoclonal antibodies were obtained from ascites induced in BALB/c nude mice by inoculation of the R4-6A2 B cell hybridoma (ATCC cell line HB 170). Rabbit
i.p.,
and media media staining
the Bovine MO).
Abbreviations: BCG, min; CFU, colony-forming
AND METHODS
culture
using
teria, Tween
Other cytokines have been characterized that are neutrophil chemoattractants. These include neutrophil-activating peptide 1 (NAP-1)/IL-8 [13] in humans and macrophage inflammatory proteins MIP-1 and MIP-2 in mice [14-16]. We have recently described the acute recruitment of massive numbers of neutrophils in mice primed by an intraperitoneal infection with Mycobacterium bovis bacille Calmette-Gu#{233}rin (BCG) following challenge with specific antigen [11. This reaction was dependent on T cells of both CD44 and CD8 phenotype [1]. Here we address the question of which mediators are involved in such mechanisms.
Bacterial hematological
#{182}4bel Salazar”
Germany). (St. Louis,
prepared
tumor
Neutrophils can be recruited by either nonspecific mechanisms or immune-mediated reactions. T cells are able to regulate neutrophil influx in response to bacteria such as in mycobacterial [1, 2] and listerial [3, 4] infections. T lymphocytes may be involved in the recruitment of neutrophils either by secreting neutrophil chemoattractants [5, 6] or by secreting cytokines able to modulate the activity of other cell types that are responsible for the direct recruitment of neutrophils. Of the many well-defined cytokines, interleukin-1 (IL-i), tumor necrosis factor (TNF), and interferon--y (IFN-y) were shown to induce the accumulation of neutrophils by indirect mechanisms involving the macrophage [7-12].
Reagents
and
M. avium medium
1992.
INTRODUCTION
MATERIALS
Cient(Jlca)
Portugal
Alegre October
Leukocyte
Rui
823,
Appelberg,
Centro
de
Citologia
Experimental,
Rua
1992
303
4100 Porto,
21,
Biology
1991;
Portugal. accepted April
Volume
52,
20,
1992.
September
8
RESULTS
CD
LU BCG-primed
mice
were
inoculated
i.p.
with
50
tg
of BCG
or M. avium antigen protein or- the same amount of BSA in 0.5 ml of phosphate-buffered saline (PBS), and the peritoneal exudate was studied 5 h later. As shown in Figure 1, both mycobacterial antigen preparations induced a massive influx of neutrophils, whereas BSA had only a small inflammatory effect. To assess whether IFN--y or TNF-a was involved in the recruitment of neutrophils in this model, primed animals were inoculated with either the recombinant cytokines or BCG antigen in the presence or absence of antibodies specific for one of the cytokines. The inoculation of 1000 U of IFN-n1 in the peritoneal cavity of normal mice evoked a small influx of neutrophils 5 h after inoculation (Fig.2). The same treatment in primed animals did not differ in magnitude, trophil
although recruitment
compared 2500 U significantly
to of
these
animals had after mycobacterial
greatly
enhanced neuantigen triggering
the control mice (Fig.2). The inoculation of recombinant TNF-a in normal mice induced (P < .05) higher neutrophil accumulation at 5
h than the administration neutrophils accumulating animals was comparable
of PBS (Fig.3). The numbers of after TNF-a treatment of primed to that of normal mice but were not significantly different from the numbers accumulating after the inoculation of PBS (Fig.3). In the same experiment, the neutrophil influx after antigen inoculation was significantly higher in primed than in control animals (P < .01). Addition of anti-TNF-a polyclonal antibodies to the antigen preparation (50 l of serum per animal) increased the response to antigen in primed animals (Fig. 3). In another experiment, anti-TNF-a serum (25 tl of serum per animal) led to a significant enhancement (P < .05) of the influx of neutrophils following antigen stimulation of primed mice, but anti-IFN-’y (50 tl of ascitic fluid per animal) had no effect on that response (Fig. 4). The role of the small inflammatory cytokines MIP-1 and MIP-2 was then assessed by looking at the effects of anti-MIP-1 or anti-MIP-2 antiserum on the antigeninduced neutrophil response in primed animals. Coinoculation of 150 tl of either serum with antigen reduced (P < .01 for both treatments) the accumulation of neutrophils at 5 h
0 1
6
a)
4
.c Q. 0 2
4-
a) z
IFN
ag.
BCG
Avium
ag.
Treatment Fig.
2.
trol,
uninfected
after gen
Number
of neutrophils mice
the inoculation or M. avium
after
(light
inoculation in animals 5). In another oculated either
there
U of IFN--y The peritoneal
reduced
to 61%, no additive
were
peritoneal
BCC-primed
cavity
mice
of con-
(heavy
shading)
or 50 g of BCG protein exudates were studied
to about half the magnitude given antigen plus nonimmune experiment, similar doses of separately or in combination
Anti-MIP-1
anti-MIP-2
in the
or
anti5 h
challenge.
after found
antigen.
shading)
of either 1000 protein antigen.
antigenic
the
accumulating
and
of the response serum (Fig. serum were intogether with influx to 73%, to 72%; thus
neutrophil
both
sera
together
effects.
DISCUSSION We previously tion of neutrophil
showed
that
influx
peritoneal
infections
described
the
T cells
with
involved
the
the
in
chronic
mycobacteria of
priming
are
the
during
stages
[2].
also BCGof neutrophils folinto the peritoneal
peritoneal
We
regulaof mouse
have
cavity
of
infected lowing
mice for enhanced recruitment the inoculation of specific antigen
cavities trophil mediated
of such animals [1]. That antigen-triggered influx was maximal at 5 h after challenge and by T cells of both CD4 and CD8 phenotypes
6
8
neuwas [1].
12
CD
uJ
10
0 6
‘C a)
6
4 .
0. 0 a-
4
4-
a)
2
z
0
PBS
8cx3
avium Antigen
Fig.
1. Number
inoculated with with 106 CFU avium cytosolic
304
of neutrophils
accumulating
of Leukocyte
Biology
ag
ag
+
antiTNF
Treatment
(50sg/animal) in the
peritoneal
cavity
of mice
50 g of protein antigen 3 weeks after they were infected i.p. of BCG. Antigenic preparations consisted of BCG or M. protein or of bovine serum albumin (BSA).
Journal
TNF
BSA
Volume
52,
September
Fig. 3. Number
of neutrophils
trol,
mice
uninfected
after
the
tein
antigen
peritoneal
1992
inoculation in the exudates
(light
accumulating shading)
of either absence were
PBS, or
studied
in the peritoneal
or BCG-primed 2500
presence
U of TNF-a, of 50 sl
5 h after
the
mice or
of anti-TNF-a antigenic
cavity (heavy
of conshading)
50 jsg of BCG antisera. challenge.
proThe
Here
we
analyzed
the
participation
of
cytokines
in
12
this
response. The response to inoculation of either recombinant IFN-n1 or TNF-a in the peritoneal cavity of BCG-primed animals did not mimic the response to antigen because it did not differ from the response in control, uninfected animals and it was of rather low magnitude compared to the response to mycobacterial antigens. These two cytokines are known to induce neutrophil recruitment by indirect mechanisms that
probably secretion
involve of the
the macrophage as the effector cell in the neutrophil chemoattractant [8-10]. Because these cytokines can be secreted by Thi cells [18] known to mediate delayed-type hypersensitivity of the Jones-Mote type with significant involvement of neutrophils [19], they were likely candidates in our model of T cell-mediated inflammation. However, the present results did not substantiate this hypothesis and even suggest that TNF-a may have an anti-inflammatory role, because the presence of anti-TNF-a specific antibodies during our in vivo assay allowed enhanced accumulation of neutrophils in response to antigen. We tested
whether
MIP-1
response
and
to
MIP-2
were
by
administering
antigen
involved
1
8
‘C U)
6
0. 0 a-
4 ‘‘‘I,, ‘‘‘S.’’
4-
*
S* ‘I,,,, 5SS*
2
a)
z
in the
upon MIP-2,
antigenic which, of
stimulation,
by
other
directly
themselves,
MIP-1
produce
recruit
neutrophils.
‘‘5’,’’
0 Cont.
we
is
unable to abolish completely the anti-MIP sera.
5.
Number
BCG-primed
mice
presence
The
neutrophils
of
of
serum
values
peritoneal
inside
exudates
were with
1. Appelberg, rophages
R.
Mycobacterial enhanced
for
by
4. Czuprinsky, bacterial
C.J., resistance
and and
accumulation H.,
0.,
and
tor
4
anti
TNF
Treatment
(50
or
d
peritoneal
of ascitic
exudates
fluid/animal),
were
studied
in
anti-IFN--y
anti-TN
5 h after
F-a
the
the
peritoneal
protein
cavity antigen
monoclonal
antisera
antigenic
(25
l/animal).
challenge.
of
in
the
antibodies The
A.B.
cells
PA. and
from
mice
Lyt2
with
1985. of antimacro-
and
listeria-immune
T
1987. J.M.,
Mrowietz,
determination activating
R.E.,
peptide
Lamb,
and
lympho-
Biochem.
(LYNAP).
1988. JR.,
Tsai,J.-J.,
neutrophil
T
U.,
of a human
883-890,
Antigen-induced human
Enhanced
immunized
Dual regulation neutrophil
Schroder,
151,
mac-
Leukoc.
macrophages
3449-3454,
and
and j
chronic neuExp. ImmunoL
Campbell,
134,
L3T4
J.,
Commun.
Cromwell,
chemotactic Immunology
lymphocytes.
fac65,
1988. C.J.,
murine
and
Brown,
interleukin-la
peritoneal
J.F.
Purified
induced
neutrophils
human
and
accumulation
and
of
mononuclear
recom-
inflam-
phagocytes:
Pathog. 3, 377-386, 1987. 8. Faccioli, L.H., Souza, G.E.P., Cunha, F.Q, Poole, S., and Ferreira, S.H. Recombinant interleukin-1 and tumor necrosis factor induce neutrophil migration “in vivo” by indirect mechanisms. Agents Actions 30, 344-349, 1990. possible
IFN
T
Structure
cloned
Czuprynski,
matory
0
and
Brown, J.F. inflammatory
P., O’Hehir,
from
binant
2
BCG
control.
challenge.
primes T cells neutrophils.
of
neutrophils
287-293,
Young,
Kay,
605-609, 7.
of
antigenic
T cell-dependent C/in.
Immunol.
neutrophil
Ret.
6. Maestrelli,
6
60, E.
derived
Biophys.
g
the
(150
of the
infections. P.M.,
by
Immunology
cyte
MT.
of
j
8
accumulating
anti-MIP-2
percentages
5 h after
inflammatory
transfer
Christophers,
with 50 (NIS),
or
the
infection recruitment
1989. C.J., Henson, of
5. Gregory,
10
neutrophils
are
mycobacterial
monocytogenes.
cells.
of
anti-MIP-l,
columns
472-477, 1992. R., and Silva, during
phage
Number
in the peritoneal cavity of BCG protein antigen in the
of
(Cont.),
studied
Listeria
12
4.
were
g
REFERENCES
trophilia
14
BCG-primed mice inoculated presence of nonimmune serum
the
50
The author is indebted to Dr. Barbara Sherry for the generous gift of antisera, to Dr. M.T Silva for helpful discussions, and to J.M. Pedrosa for skillful assistance. This work was supported by funds from the JNICT and INIC (Lisbon) and the Damien Foundation (Brussels).
mediated
Fig.
anti-MIP2
ACKNOWLEDGMENTS
accumulation
anti
with
nonimmune The
accumulating
inoculated
78, 478-483, 3. Czuprinsky,
NIS
anti-MIP1
Treatment Fig.
Biol. 51, 2. Appelberg,
In-
also possible, since the neutrophil recruitment
cytokines
*
‘S.’’’’
specific
polyclonal antisera together with the antigen. We showed that both cytokines were involved because there was a significant reduction in the neutrophil accumulation triggered by antigen at the time point studied. When both antisera were used in combination, no additive effects were seen, suggesting that the two cytokines may act in different but consecutive steps of the cascade of events that lead to neutrophil recruitment and accumulation in the peritoneal cavity. This hypothesis is supported by the fact that only MIP-2 is highly chemotactic for neutrophils in vitro. MIP-1 may thus act in a way other than chemotaxis (e.g., modulation of cell adhesion, vascular changes). We postulate that T
volvement
10
0
tl/animal).
neutrophilic
cells, and
(0 LU
contributions
9. Harmsen, AG., and macrophages of
neutrophils
297-302,
Appelberg
to antibacterial
and in in
Havell, E.A. Roles lipopolysaccharide-induced
cutaneous
air
Microb.
resistance.
pouches.
of tumor
necrosis factor accumulation
Infect.
Immun.
58,
1990.
MIP-induced
neutrophil
recruitment
305
10. Ribeiro, R.A., Cunha, F.Q., and gamma interferon causes neutrophil release of a macrophage neutrophil Exp. Pathol. 71, 717-725, 1990. 11.
Sayers,
T.J.,
Wiltrout,
AM., and neutrophil
T.A.,
Bull,
Ferreira, S.H. Recombinant migration mediated by the chemotactic factor. mt. j C.A.,
induction
polymorphonuclear 13.
14.
Leukoc. Leonard,
E.J.,
tion
protein-i
Biol.
2, 479,
Davatelis, Luedke, ami, A.
Denn,
A.C.,
III,
Pilaro,
of
infiltration
by
IL-i
migration
43,
349-356,
1988.
and
Yoshimura,
T
(NAP-I
[interleukin-8]).
during
Neutrophil
Am.
and
inflammation.
Cell MoL
Respir.
Journal
16.
18.
inflammatory
of Leukocyte
S.D.,
protein
Biology
(MIP),
a novel
Volume
52,
September
2:
cDNAs for its human Cerami,
A.
members
of
3, 2565-2573,
Mosmann,
T.R., patterns
tional
properties.
Fong,
TAT.,
murine macrophage homologues. j Exp. Macrophage
a
novel
Exp.
Med.
inflamma-
Med.
172,
inflammatory
superfamily
and
Coffman,
Annu. and
Rev.
Mosmann, mediated
1989.
pro-
of
cytokines.
1989.
R.L.
of lymphokine
hypersensitivity
2887-2893,
1992
Deventer,
Silva, MT., Silva, M.N.T., and Appelberg, macrophage cooperation in the host defense terial infections. Microb. Pathog. 6, 369-380,
type
monokine
and
I and
different
19.
j
properties.
C., Bauer, D., McClain, J., Sherry, S., and Cerami, A. Cloning and
P., Gallegos, van
of 2 and
FASEBJ 17.
chemokinetic
1990.
Wolpe,
teins
1990.
J.,
and
1988.
M.,
characterization tory protein
attractant/activa-
J.
Fabre,
911-919,
G., Tekamp-olson, P., Wolpe, S.D., Hermsen, K., C., Gallegos, C., Coit, D., Merryweather, and CerCloning and characterization of a cDNA for murine
macrophage
306
leukocyte
BioL
1939-1944,
B.,
tumor necrosis factor. j Immunol. 141, 1670-1677, 1988. 12. Wankowicz, Z., Megyeri, P., and Issekutz, A. Synergy between tumour necrosis factor a and interleukin-1 in the induction of j
inflammatory
167,
15. Tekamp-Olson,
Lokesh, B. Effect of cytokines on polymorphonuclear infiltration in the mouse. Prostaglandin-and
leukotriene-independent
with
R. against 1989.
TH1
secretion
and
lead
to
TH2 different
Immunol.
7,
T.R.
The
role
of IFNy
Thl
clones.
j
by
Neutrophilmycobac-
145-173,
cells: func-
1989.
in delayed-
ImmunoL
143,
proteins MIP-1 and MIP-2 neutrophil recruitment
are
Rui Appelberg Centro
de Cilologia
Institute,
Abstract:
Mice
Mycobacterium
University
Experimental
(Instituto
of Porto,
Nacional
de Investigacdo
intraperitoneally
(i.p.)
infected
with
formed
bovis
bacille Calmette-Gu#{233}rin (BCG) respond to an i.p. challenge with mycobacterial antigen with an acute and extensive accumulation of neutrophils. This influx was not mimicked by the inoculation of recombinant interferon--y (IFN-’y) or tumor necrosis factor a (TNF-a). The antigen-induced recruitment of neutrophils was not affected by coinoculation of anti-IFN-y antibodies, was enhanced by anti-TNF-a antisera, and was significantly reduced by antisera against macrophage inflammatory proteins MIP-1 and MIP-2. The latter two sera had no additive effects. We hypothesize that mycobacteria-specific T cells are triggered by antigen to secrete MIP-1 and MIP-2, which directly mediate, at least partly, the influx of neutrophils that takes place in
J.
this
model.
Key
Words: factor
necrosis
Leukoc. neutrophils, macrophage
.
Biol.
52:
303-306;
T cell-mediated inflammatory
. interferon-n1 proteins
Hemacolor
macrophage
inflammatory
a generous New York).
gift
MIP-2
were
University,
Bacteria
Sciences
(Detroit,
preparations
from
Merck
(Darmstadt,
and
bacterial
from
antigen
MIP-1 and (Rockefeller
proteins
Dr.
B. Sherry
preparations
bovis BCG (TMCC 1011, strain Pasteur) and (ATCC 25291) were grown in Middlebrook 7H9 supplemented with 0.04% Tween 80. Inocula were from mid-log phase cultures by pelleting the bacresuspending in a small volume of saline with 0.04% 80, briefly sonicating the bacterial suspension, and
Mycobacterium
freezing at -70#{176}C until use. Listeria monocytogenes (strain EGD) was grown in Antibiotic 3 broth. Cytosolic proteins from M. bovis BCG, M. avium, and L. monocytogenes were prepared as described [1] and used as the antigen sources.
Animals Female C57BL/6 Institute (Oeiras, hygiene conditions, ad libitum.
Priming
and
mice
were Portugal), and fed
challenge
MI).
The
was
per-
intraperitoneal; necrosis
Reprint
from under and
the
Gulbenkian conventional acidified water
of mice
Received
of
bacille
Calmette-Gu#{233}rin; unit;
MIP-I,
IFN-y,
macrophage
peptide
1;
BSA,
interferon-7;
serum
IL-I,
inflammatory
PBS,
bovine protein
phosphate-buffered
albu-
interleukin-I; I;
saline;
NAP-i,
TNF,
factor.
requests:
do Campo
Journal
purchased kept food pellets
Mice were inoculated intraperitoneally (i.p.) with 106 colonyforming units (CFU) of BCG and 3 weeks later were administered the antigen or cytokine stimulus by the same route. Five hours after the second inoculation, the animals were sacrificed and the peritoneal exudate studied as described elsewhere [1, 17]. The results are expressed as the arithmetic means of neutrophils per peritoneal cavity from four mice per time point ± 1 standard deviation of the mean. The statistical analysis was done with the Statworks program using Student’s I-test.
tumor
Difco
kit
antimouse
neutrophil-activating
were from of cytospin
Biomedical
serum albumin (BSA) was from Sigma Recombinant mouse IFN-’y, recombinant mouse TNF-a, and rabbit antimouse TNF-a polyclonal serum were purchased from Genzyme (Boston, MA). Antimouse IFN--y monoclonal antibodies were obtained from ascites induced in BALB/c nude mice by inoculation of the R4-6A2 B cell hybridoma (ATCC cell line HB 170). Rabbit
i.p.,
and media media staining
the Bovine MO).
Abbreviations: BCG, min; CFU, colony-forming
AND METHODS
culture
using
teria, Tween
Other cytokines have been characterized that are neutrophil chemoattractants. These include neutrophil-activating peptide 1 (NAP-1)/IL-8 [13] in humans and macrophage inflammatory proteins MIP-1 and MIP-2 in mice [14-16]. We have recently described the acute recruitment of massive numbers of neutrophils in mice primed by an intraperitoneal infection with Mycobacterium bovis bacille Calmette-Gu#{233}rin (BCG) following challenge with specific antigen [11. This reaction was dependent on T cells of both CD44 and CD8 phenotype [1]. Here we address the question of which mediators are involved in such mechanisms.
Bacterial hematological
#{182}4bel Salazar”
Germany). (St. Louis,
prepared
tumor
Neutrophils can be recruited by either nonspecific mechanisms or immune-mediated reactions. T cells are able to regulate neutrophil influx in response to bacteria such as in mycobacterial [1, 2] and listerial [3, 4] infections. T lymphocytes may be involved in the recruitment of neutrophils either by secreting neutrophil chemoattractants [5, 6] or by secreting cytokines able to modulate the activity of other cell types that are responsible for the direct recruitment of neutrophils. Of the many well-defined cytokines, interleukin-1 (IL-i), tumor necrosis factor (TNF), and interferon--y (IFN-y) were shown to induce the accumulation of neutrophils by indirect mechanisms involving the macrophage [7-12].
Reagents
and
M. avium medium
1992.
INTRODUCTION
MATERIALS
Cient(Jlca)
Portugal
Alegre October
Leukocyte
Rui
823,
Appelberg,
Centro
de
Citologia
Experimental,
Rua
1992
303
4100 Porto,
21,
Biology
1991;
Portugal. accepted April
Volume
52,
20,
1992.
September
8
RESULTS
CD
LU BCG-primed
mice
were
inoculated
i.p.
with
50
tg
of BCG
or M. avium antigen protein or- the same amount of BSA in 0.5 ml of phosphate-buffered saline (PBS), and the peritoneal exudate was studied 5 h later. As shown in Figure 1, both mycobacterial antigen preparations induced a massive influx of neutrophils, whereas BSA had only a small inflammatory effect. To assess whether IFN--y or TNF-a was involved in the recruitment of neutrophils in this model, primed animals were inoculated with either the recombinant cytokines or BCG antigen in the presence or absence of antibodies specific for one of the cytokines. The inoculation of 1000 U of IFN-n1 in the peritoneal cavity of normal mice evoked a small influx of neutrophils 5 h after inoculation (Fig.2). The same treatment in primed animals did not differ in magnitude, trophil
although recruitment
compared 2500 U significantly
to of
these
animals had after mycobacterial
greatly
enhanced neuantigen triggering
the control mice (Fig.2). The inoculation of recombinant TNF-a in normal mice induced (P < .05) higher neutrophil accumulation at 5
h than the administration neutrophils accumulating animals was comparable
of PBS (Fig.3). The numbers of after TNF-a treatment of primed to that of normal mice but were not significantly different from the numbers accumulating after the inoculation of PBS (Fig.3). In the same experiment, the neutrophil influx after antigen inoculation was significantly higher in primed than in control animals (P < .01). Addition of anti-TNF-a polyclonal antibodies to the antigen preparation (50 l of serum per animal) increased the response to antigen in primed animals (Fig. 3). In another experiment, anti-TNF-a serum (25 tl of serum per animal) led to a significant enhancement (P < .05) of the influx of neutrophils following antigen stimulation of primed mice, but anti-IFN-’y (50 tl of ascitic fluid per animal) had no effect on that response (Fig. 4). The role of the small inflammatory cytokines MIP-1 and MIP-2 was then assessed by looking at the effects of anti-MIP-1 or anti-MIP-2 antiserum on the antigeninduced neutrophil response in primed animals. Coinoculation of 150 tl of either serum with antigen reduced (P < .01 for both treatments) the accumulation of neutrophils at 5 h
0 1
6
a)
4
.c Q. 0 2
4-
a) z
IFN
ag.
BCG
Avium
ag.
Treatment Fig.
2.
trol,
uninfected
after gen
Number
of neutrophils mice
the inoculation or M. avium
after
(light
inoculation in animals 5). In another oculated either
there
U of IFN--y The peritoneal
reduced
to 61%, no additive
were
peritoneal
BCC-primed
cavity
mice
of con-
(heavy
shading)
or 50 g of BCG protein exudates were studied
to about half the magnitude given antigen plus nonimmune experiment, similar doses of separately or in combination
Anti-MIP-1
anti-MIP-2
in the
or
anti5 h
challenge.
after found
antigen.
shading)
of either 1000 protein antigen.
antigenic
the
accumulating
and
of the response serum (Fig. serum were intogether with influx to 73%, to 72%; thus
neutrophil
both
sera
together
effects.
DISCUSSION We previously tion of neutrophil
showed
that
influx
peritoneal
infections
described
the
T cells
with
involved
the
the
in
chronic
mycobacteria of
priming
are
the
during
stages
[2].
also BCGof neutrophils folinto the peritoneal
peritoneal
We
regulaof mouse
have
cavity
of
infected lowing
mice for enhanced recruitment the inoculation of specific antigen
cavities trophil mediated
of such animals [1]. That antigen-triggered influx was maximal at 5 h after challenge and by T cells of both CD4 and CD8 phenotypes
6
8
neuwas [1].
12
CD
uJ
10
0 6
‘C a)
6
4 .
0. 0 a-
4
4-
a)
2
z
0
PBS
8cx3
avium Antigen
Fig.
1. Number
inoculated with with 106 CFU avium cytosolic
304
of neutrophils
accumulating
of Leukocyte
Biology
ag
ag
+
antiTNF
Treatment
(50sg/animal) in the
peritoneal
cavity
of mice
50 g of protein antigen 3 weeks after they were infected i.p. of BCG. Antigenic preparations consisted of BCG or M. protein or of bovine serum albumin (BSA).
Journal
TNF
BSA
Volume
52,
September
Fig. 3. Number
of neutrophils
trol,
mice
uninfected
after
the
tein
antigen
peritoneal
1992
inoculation in the exudates
(light
accumulating shading)
of either absence were
PBS, or
studied
in the peritoneal
or BCG-primed 2500
presence
U of TNF-a, of 50 sl
5 h after
the
mice or
of anti-TNF-a antigenic
cavity (heavy
of conshading)
50 jsg of BCG antisera. challenge.
proThe
Here
we
analyzed
the
participation
of
cytokines
in
12
this
response. The response to inoculation of either recombinant IFN-n1 or TNF-a in the peritoneal cavity of BCG-primed animals did not mimic the response to antigen because it did not differ from the response in control, uninfected animals and it was of rather low magnitude compared to the response to mycobacterial antigens. These two cytokines are known to induce neutrophil recruitment by indirect mechanisms that
probably secretion
involve of the
the macrophage as the effector cell in the neutrophil chemoattractant [8-10]. Because these cytokines can be secreted by Thi cells [18] known to mediate delayed-type hypersensitivity of the Jones-Mote type with significant involvement of neutrophils [19], they were likely candidates in our model of T cell-mediated inflammation. However, the present results did not substantiate this hypothesis and even suggest that TNF-a may have an anti-inflammatory role, because the presence of anti-TNF-a specific antibodies during our in vivo assay allowed enhanced accumulation of neutrophils in response to antigen. We tested
whether
MIP-1
response
and
to
MIP-2
were
by
administering
antigen
involved
1
8
‘C U)
6
0. 0 a-
4 ‘‘‘I,, ‘‘‘S.’’
4-
*
S* ‘I,,,, 5SS*
2
a)
z
in the
upon MIP-2,
antigenic which, of
stimulation,
by
other
directly
themselves,
MIP-1
produce
recruit
neutrophils.
‘‘5’,’’
0 Cont.
we
is
unable to abolish completely the anti-MIP sera.
5.
Number
BCG-primed
mice
presence
The
neutrophils
of
of
serum
values
peritoneal
inside
exudates
were with
1. Appelberg, rophages
R.
Mycobacterial enhanced
for
by
4. Czuprinsky, bacterial
C.J., resistance
and and
accumulation H.,
0.,
and
tor
4
anti
TNF
Treatment
(50
or
d
peritoneal
of ascitic
exudates
fluid/animal),
were
studied
in
anti-IFN--y
anti-TN
5 h after
F-a
the
the
peritoneal
protein
cavity antigen
monoclonal
antisera
antigenic
(25
l/animal).
challenge.
of
in
the
antibodies The
A.B.
cells
PA. and
from
mice
Lyt2
with
1985. of antimacro-
and
listeria-immune
T
1987. J.M.,
Mrowietz,
determination activating
R.E.,
peptide
Lamb,
and
lympho-
Biochem.
(LYNAP).
1988. JR.,
Tsai,J.-J.,
neutrophil
T
U.,
of a human
883-890,
Antigen-induced human
Enhanced
immunized
Dual regulation neutrophil
Schroder,
151,
mac-
Leukoc.
macrophages
3449-3454,
and
and j
chronic neuExp. ImmunoL
Campbell,
134,
L3T4
J.,
Commun.
Cromwell,
chemotactic Immunology
lymphocytes.
fac65,
1988. C.J.,
murine
and
Brown,
interleukin-la
peritoneal
J.F.
Purified
induced
neutrophils
human
and
accumulation
and
of
mononuclear
recom-
inflam-
phagocytes:
Pathog. 3, 377-386, 1987. 8. Faccioli, L.H., Souza, G.E.P., Cunha, F.Q, Poole, S., and Ferreira, S.H. Recombinant interleukin-1 and tumor necrosis factor induce neutrophil migration “in vivo” by indirect mechanisms. Agents Actions 30, 344-349, 1990. possible
IFN
T
Structure
cloned
Czuprynski,
matory
0
and
Brown, J.F. inflammatory
P., O’Hehir,
from
binant
2
BCG
control.
challenge.
primes T cells neutrophils.
of
neutrophils
287-293,
Young,
Kay,
605-609, 7.
of
antigenic
T cell-dependent C/in.
Immunol.
neutrophil
Ret.
6. Maestrelli,
6
60, E.
derived
Biophys.
g
the
(150
of the
infections. P.M.,
by
Immunology
cyte
MT.
of
j
8
accumulating
anti-MIP-2
percentages
5 h after
inflammatory
transfer
Christophers,
with 50 (NIS),
or
the
infection recruitment
1989. C.J., Henson, of
5. Gregory,
10
neutrophils
are
mycobacterial
monocytogenes.
cells.
of
anti-MIP-l,
columns
472-477, 1992. R., and Silva, during
phage
Number
in the peritoneal cavity of BCG protein antigen in the
of
(Cont.),
studied
Listeria
12
4.
were
g
REFERENCES
trophilia
14
BCG-primed mice inoculated presence of nonimmune serum
the
50
The author is indebted to Dr. Barbara Sherry for the generous gift of antisera, to Dr. M.T Silva for helpful discussions, and to J.M. Pedrosa for skillful assistance. This work was supported by funds from the JNICT and INIC (Lisbon) and the Damien Foundation (Brussels).
mediated
Fig.
anti-MIP2
ACKNOWLEDGMENTS
accumulation
anti
with
nonimmune The
accumulating
inoculated
78, 478-483, 3. Czuprinsky,
NIS
anti-MIP1
Treatment Fig.
Biol. 51, 2. Appelberg,
In-
also possible, since the neutrophil recruitment
cytokines
*
‘S.’’’’
specific
polyclonal antisera together with the antigen. We showed that both cytokines were involved because there was a significant reduction in the neutrophil accumulation triggered by antigen at the time point studied. When both antisera were used in combination, no additive effects were seen, suggesting that the two cytokines may act in different but consecutive steps of the cascade of events that lead to neutrophil recruitment and accumulation in the peritoneal cavity. This hypothesis is supported by the fact that only MIP-2 is highly chemotactic for neutrophils in vitro. MIP-1 may thus act in a way other than chemotaxis (e.g., modulation of cell adhesion, vascular changes). We postulate that T
volvement
10
0
tl/animal).
neutrophilic
cells, and
(0 LU
contributions
9. Harmsen, AG., and macrophages of
neutrophils
297-302,
Appelberg
to antibacterial
and in in
Havell, E.A. Roles lipopolysaccharide-induced
cutaneous
air
Microb.
resistance.
pouches.
of tumor
necrosis factor accumulation
Infect.
Immun.
58,
1990.
MIP-induced
neutrophil
recruitment
305
10. Ribeiro, R.A., Cunha, F.Q., and gamma interferon causes neutrophil release of a macrophage neutrophil Exp. Pathol. 71, 717-725, 1990. 11.
Sayers,
T.J.,
Wiltrout,
AM., and neutrophil
T.A.,
Bull,
Ferreira, S.H. Recombinant migration mediated by the chemotactic factor. mt. j C.A.,
induction
polymorphonuclear 13.
14.
Leukoc. Leonard,
E.J.,
tion
protein-i
Biol.
2, 479,
Davatelis, Luedke, ami, A.
Denn,
A.C.,
III,
Pilaro,
of
infiltration
by
IL-i
migration
43,
349-356,
1988.
and
Yoshimura,
T
(NAP-I
[interleukin-8]).
during
Neutrophil
Am.
and
inflammation.
Cell MoL
Respir.
Journal
16.
18.
inflammatory
of Leukocyte
S.D.,
protein
Biology
(MIP),
a novel
Volume
52,
September
2:
cDNAs for its human Cerami,
A.
members
of
3, 2565-2573,
Mosmann,
T.R., patterns
tional
properties.
Fong,
TAT.,
murine macrophage homologues. j Exp. Macrophage
a
novel
Exp.
Med.
inflamma-
Med.
172,
inflammatory
superfamily
and
Coffman,
Annu. and
Rev.
Mosmann, mediated
1989.
pro-
of
cytokines.
1989.
R.L.
of lymphokine
hypersensitivity
2887-2893,
1992
Deventer,
Silva, MT., Silva, M.N.T., and Appelberg, macrophage cooperation in the host defense terial infections. Microb. Pathog. 6, 369-380,
type
monokine
and
I and
different
19.
j
properties.
C., Bauer, D., McClain, J., Sherry, S., and Cerami, A. Cloning and
P., Gallegos, van
of 2 and
FASEBJ 17.
chemokinetic
1990.
Wolpe,
teins
1990.
J.,
and
1988.
M.,
characterization tory protein
attractant/activa-
J.
Fabre,
911-919,
G., Tekamp-olson, P., Wolpe, S.D., Hermsen, K., C., Gallegos, C., Coit, D., Merryweather, and CerCloning and characterization of a cDNA for murine
macrophage
306
leukocyte
BioL
1939-1944,
B.,
tumor necrosis factor. j Immunol. 141, 1670-1677, 1988. 12. Wankowicz, Z., Megyeri, P., and Issekutz, A. Synergy between tumour necrosis factor a and interleukin-1 in the induction of j
inflammatory
167,
15. Tekamp-Olson,
Lokesh, B. Effect of cytokines on polymorphonuclear infiltration in the mouse. Prostaglandin-and
leukotriene-independent
with
R. against 1989.
TH1
secretion
and
lead
to
TH2 different
Immunol.
7,
T.R.
The
role
of IFNy
Thl
clones.
j
by
Neutrophilmycobac-
145-173,
cells: func-
1989.
in delayed-
ImmunoL
143,
